2011-07-05
Actions from previous meeting
Actions from this meeting
Weekly Summary
Default umask is now 002 for private user groups, for easier data sharing. (umask-to-0002) (Martin)
- Thunderbird replaced Evolution as default email client on the CDs. (Chris Coulson)
- Apport is now enabled by default again for Alpha 2. (Martin)
- The previous item made the CDs explode again (~ 15 MB oversized), but we should now have all big new Oneiric things on the CDs.
Good progress on OneConf, server side will be helped by mvo and stgraber. API defined, will work on that while offline (on train) next Friday.
Partner update
Unity
- New unity snapshot with new nux/unity (and gtk3 indicator support).
- Compiz patch was needed with the new unity dialogs
- but this brings a lot of issue with compiz current configuration handling system (can't add new plugin, even when existing plugins depends on a new one, making everyone segfaulting). Need to prioritize the gsettings backend and remove those limitation in libcompizconfig itself.
- Unity-2d new release with some metacity patch and a lot of bug fixes.
- Unity 2d is now at 80% accessible. Some patches post alpha2 in Qt will help removing the latest gotchas. Very active work with upstream to get that done as soon as possible.
- Note that the half implement unity-like modal dialogs showing in metacity due to the metacity theme used in compiz will be removed (no ressource for implementing the full design).
- Still using gtk2 indicators, but good progress (merge request today in trunk) to use the unity-panel service and so, get rid of it.
- Qt has now the sni-qt support (seeded by default, will be there after alpha2): transparent indicator fallback support. This mean no more systray for popular Qt apps like mumble, skype!
- Incoming dconf support in QML (and not Qt). Will be used by unity 2d first
- Discussion with upstream with zeitgeist first indexing to not have blank unity places. Still need to determine if this should be integrated (as it is today) with the activity log manager managing blacklists as well, if so, if it should be added to gnome-control-center and such or just run on first login. Needs design team to be back from holidays to discuss that. Add items to alpha3
Software Center
- Released Software Center version 4.1.7 and 4.1.8:
- Features UI enhancements and many bug fixes
- Adds support in the client for removing and modifying reviews, many thanks to Aaron Peachey!
- Awaits server-side support to enable the functionality, work is in progress
GTK3 porting work continues, Matthew McGowan's branch has been merged to trunk now
- Good progress in the version 5.0 redesign work by Canonical UX team
- Finalized design and mockups in progress
Kubuntu
X.org
- The first mesa 7.11 snapshot package has been uploaded. Nothing appears to have caught fire
- No problems were raised at the nouveau 3D testing day in Dublin, and positive results were reported.
IRC log Western edition
17:35:11 kenvandine pitti, indicator-power is ready to be packaged! 17:35:17 kenvandine i'll get on that real soon 17:35:31 pitti kenvandine: I'm confused about this a bit -- isn't it going to move into gnome-settings-daemon? 17:35:39 kenvandine pitti, and i never heard how your meeting went about the U1 release stuff 17:35:41 pitti so we won't use that after all, and instead use indicator-power? 17:35:46 seb128 hum 17:35:53 seb128 where is the partner update? 17:35:54 pitti what do we do in the classic session then? 17:36:02 pitti seb128: I mis-looked, ignore me 17:36:11 seb128 ok 17:36:23 kenvandine seb128, what will happen for power in the classic session? 17:36:38 seb128 it will have no icon or indicator 17:36:39 * kenvandine doesn't really know... but for unity we need one 17:37:02 pitti not shipping the g-s-d plugin sounds a bit bad to me -- can it live alongside indicator-power somehow? 17:37:07 seb128 pitti, what moves in g-s-d is the service 17:37:22 seb128 the icon is dropped in favor of a gnome-shell indicator 17:37:27 pitti ah 17:37:30 seb128 or an system indicator in our case 17:37:35 pitti so i-power will actually talk to the g-s-d plugin? 17:37:53 rodrigo_ pitti, yes, the indicator would use the g-s-d plugin 17:37:53 seb128 or upower directly 17:38:03 seb128 i.e dbus calls to get datas 17:38:03 pitti kenvandine: U1 release> I had a quick chat with Jason, but we didn't actually have a meeting about it which included me 17:38:03 rodrigo_ just needs to call the new interface, which has changed name 17:38:14 pitti ah, great 17:38:24 seb128 pitti, the indicator is just indicating infos it collects over dbus 17:38:27 kenvandine pitti, oh... ok, so we still don't know what the plan is... 17:38:33 * kenvandine will get on them 17:40:34 kenvandine that is all i have 17:40:38 pitti thanks 17:41:18 pitti didrocks: so, unity 17:41:24 pitti thanks for the wiki report 17:41:28 didrocks all is on the wiki, as usual :) 17:41:37 pitti didrocks: I'm a bit confused about the compizconfig comment 17:42:02 didrocks ok, so, basically, this is an issue with configuration management in compiz 17:42:18 pitti did that change recently? 17:42:22 didrocks like, you can't add new plugins to existing user (like unitydialog), even if they never touched their current settings 17:42:43 didrocks no, it's just that I did some unity --reset automatically at some point of the release last cycle 17:42:47 pitti aah 17:42:56 didrocks but now, unity depends on the gtk_init plugin 17:42:56 pitti didrocks: so we actually did have this problem in natty as well, it just didn't hurt us yet? 17:43:05 didrocks and compiz isn't able to add the gtk_init automatically 17:43:13 didrocks if we don't patch compiz, everyone will segfault 17:43:16 pitti i. e. compiz plugins can't have dependencies? 17:43:19 didrocks pitti: right 17:43:28 didrocks they can, but they are only resolved in ccsm 17:43:32 didrocks by the graphical interface 17:43:53 didrocks which is kind of weird for a project based on plugin, but well :) 17:44:13 didrocks so yeah, I try to convince dx to reshuffle their priorities for that 17:44:27 didrocks right now, the working is forcing compiz to call gtk_init in main() (for everyone then) 17:44:41 didrocks but that only works because we are the only one to use it and only in that case 17:45:00 didrocks will be nice in the future to be able to change the default plugin list if the user didn't touch it 17:46:00 pitti didrocks: the unity plugin can't call gtk_init()? 17:46:05 pitti I thought they were all in-process 17:46:14 didrocks pitti: right, but all plugins are dlopened 17:46:25 didrocks and now, unity and unity-dialogs are using gtk 17:46:31 didrocks and gtk_init doesn't support this case 17:46:43 pitti gtk_init() doesn't work from a dlopen()'ed lib? 17:46:55 didrocks if more than once are doing that, it results in a hang 17:47:11 didrocks (previous upstream tarball, hence the dialogs not being there before) 17:47:30 pitti didrocks: ok, at this point I think we should move the discussion to a bug report 17:47:53 didrocks pitti: we discussed about it with desrt at the sprint, it seems it's not something they want to support 17:48:03 didrocks but yeah, can be moved in a bug report if you prefer 17:48:44 pitti more time to ponder, and not 10 other people waiting in the meeting 17:49:31 pitti didrocks: nice progress on a11y! 17:49:49 didrocks yeah, was quite straightforward once the support was there :) 17:50:13 pitti tremolux: thanks for the s-c report; and great to read that gtk3 is now in trunk! 17:50:28 pitti tremolux: anything we need to discuss there? still blocked on design feedback? 17:51:04 tremolux pitti: yes, we don't have the full design yet, but we are planning it in stages best we can 17:51:30 tremolux pitti: juggling priorities gtk3/new design, etc. 17:53:21 pitti tremolux: bug 432552 is the remaining alpha-2 WI for the s-c blueprint 17:53:22 ubot2 Launchpad bug 432552 in software-center "USC needs clearer/simpler icon" [Medium,Confirmed] https://launchpad.net/bugs/432552 17:53:38 pitti I figure it can easily move to a3, or do we have some pressure to squeeze that into a2? 17:53:42 tremolux pitti: ah, thanks, I should have moved that 17:53:58 pitti ok -> a3 now 17:54:03 tremolux pitti: no need to squeeze that to a2, it's coming 17:54:04 tremolux thanks! 17:54:29 pitti which brings us down to 7 WIs for a2 17:54:37 pitti https://blueprints.launchpad.net/ubuntu/+spec/desktop-o-default-email-client still has 4 left 17:55:03 pitti but 3 of which are actually the same: seb128, chrisccoulson, allison to discuss strategy for creating new Ubuntu One address book (on new install) 17:55:16 pitti as this says "discuss", I figure it can stay on a2? 17:55:38 pitti it sounds as it might be a prerequisite for further upstream development 17:56:19 pitti chrisccoulson: and John still has "Review progress in styling improvements, provide guidance on further improvements needed" -> is that a blocker and should we prod him, or move to a3? 17:56:33 seb128 pitti, john in on vac this week so a3 I guess 17:57:04 seb128 dunno for the upstream discussion, I think chrisccoulson said that m_conley figured it out 17:57:09 seb128 but he said he would check 17:57:20 chrisccoulson yeah, i'll ask 17:57:36 pitti ok, we can bring this up in tomorrow's IRC meeting 17:57:40 seb128 right 17:58:36 pitti then the remaining one is on https://blueprints.launchpad.net/ubuntu/+spec/desktop-o-gwibber-gtk3 17:58:42 pitti [ken-vandine] move dupe message detection to the backend: TODO 17:58:53 pitti doesn't sound like a blocker to me, but can happen in upstream trunk, I suppose? 17:59:28 kenvandine yeah 17:59:39 kenvandine that is really just blocked on design... sort of 17:59:52 kenvandine njpatel hasn't decided how we will present them, so i am holding off 18:00:09 kenvandine it will be quick to do when i know what to do :) 18:00:25 kenvandine pitti, i am going to upload the new gwibber right after a2 18:00:38 pitti ok, so we can just move that 18:00:39 pitti thanks! 18:00:41 kenvandine yeah 18:00:44 pitti anything else we need to discuss? 18:00:50 kenvandine nope 18:01:11 pitti kenvandine: oh, that's for "[njpatel] create update entry widget: TODO"? 18:01:17 pitti kenvandine: I was talking about the dupe detection 18:01:37 kenvandine i know 18:01:50 kenvandine i just merged his entry, so i'll mark that done 18:02:17 kenvandine unrelated items 18:02:19 pitti seb128: anythign to announce or discuss for gnome 3.1? 18:02:37 seb128 not especially 18:02:58 seb128 just "claim updates, note those on the etherpad, be careful about the freeze" 18:02:58 seb128 ;-) 18:03:06 pitti deal! 18:03:09 seb128 joke aside there is quite some outdated lines on version 18:03:26 seb128 so if anyone feels like claiming some updates this week please do 18:03:58 rodrigo_ seb128, ok, will do some in between g-c-c fixes :) 18:04:18 seb128 oh, yeah, if you find any g-c-c bug assign those to rodrigo_ as well 18:04:23 seb128 he's collecting them 18:04:24 seb128 ;-) 18:04:37 pitti so, meeting adjourned; thanks everyone! 18:04:42 seb128 thanks 18:04:45 kenvandine thanks 18:04:53 mterry bye! 18:04:55 seb128 ok, my a2 install just finished 18:04:57 Sweetshark pitti: let me just append the LO status 18:05:04 pitti ah, please go ahead 18:05:07 tremolux thanks everybody 18:05:07 seb128 wouou it worked 18:05:07 Sweetshark also upstream LO broke small symbols (-g1), which currently leads to a 2GB -dbg package. It was cooperatively debuged with _rene_ from debian. This is intended to be included in debian tonight. As this was pretty much the last thing that blocked the debian 3.4.X release, we hopefully can base on that without making a 3.4.1-0ubuntu1 release. 18:05:15 rodrigo_ seb128, hmm, I should have knocked more on your door, so that you couldn't sleep :) 18:05:19 Sweetshark libreoffice status: 3.3.3-1ubuntu1 is ready for SRUing, 3.3.3-2 to 3.3.3-4 includes debian changes we do not want in natty (requires new javahelper version). libreoffice-l10n-3.3.3-1ubuntu1 still has some open merge issues (work in progress). 3.4.0 status: I got a build finished on the weekend (without mono and kde and opensymbol for now). I am currently merging in debian changes which a) fix the mono, kde, opensymbol problems b) update to 18:05:23 seb128 rodrigo_, ;-) 18:05:47 Sweetshark also upstream LO broke small symbols (-g1), which currently leads to a 2GB -dbg package. It was cooperatively debuged with _rene_ from debian. This is intended to be included in debian tonight. As this was pretty much the last thing that blocked the debian 3.4.X release, we hopefully can base on that without making a 3.4.1-0ubuntu1 release and without postponing WIs. 18:06:37 pitti Sweetshark: sounds like good progress! so I guess it's sponsoring time soon? 18:07:34 didrocks reboot, brb 18:08:10 Sweetshark pitti: for 3.3.3? yes. I think I have fixed the -l10n merge issue, build is running currently.
IRC log Eastern edition
DesktopTeam/Meeting/2011-07-05 (last edited 2011-07-06 00:53:35 by raof)